Is conditionally automated driving a bad idea? Observations from an on-road study in automated vehicles with multiple levels of driving automation
Soyeon Kim (TU Delft - Traffic Systems Engineering)
Fjollë Novakazi (Chalmers University of Technology)
I. C.Mari Anne Karlsson (Chalmers University of Technology)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
With the increasing adoption of driving automation technologies, vehicles equipped with SAE Level 3 driving automation are becoming available on the market. This study explores drivers’ behaviour when driving conditionally automated vehicles on-road, providing multiple levels of driving automation. Sixteen participants drove a Wizard-of-Oz vehicle offering several levels of automation (Manual, SAE Level 2 and Level 3) on a public highway. Data was collected during driving sessions (observations and think-aloud) and post-driving sessions (in-depth interviews). The results indicate that drivers show errors in mode transitions and mode awareness. These errors include unintended deactivation of Level 2 driving automation, confusion about driving modes after disengaging Level 3 driving automation, and confusion about the current driving mode. These findings highlight a fundamental limitation in the design of automation systems when humans are required to operate multiple modes within a single system, making it challenging to distinguish between them clearly. This ambiguity and lack of understanding affected how drivers interacted with, interpreted, and responded to the automated vehicle. The study provides insights for designing automated vehicles with multiple levels of driving automation, aiming to improve mode awareness and overall safety.